NefMoto

Hey guys,

I need some help here.  I'm working with my AMB 1.8T running a GT2871R on Eurodyne Maestro 7 software.  I am increasing the boost level past the limits of the stock MAP sensor.  However, when doing so, The ECU isn't adding the correct amount of fuel into the cylinders because it doesn't think the boost is anywhere past 22 PSI.  Fueling is fantastic below 22psi.....not so much after.

Now I know I'm certainly not the first and only person running more boost than the MAP sensor will allow, so my question is......How are you guys adding a sufficient amount of fuel past the MAP sensor limit to keep O2 corrections reasonable?

Fueling has nothing to do with MAP Sensor readings unless you're running MAFless tune. IF this is the case, there is no workaround since ECU cannot properly calculate amount of air entering the engine. Case closed.

If you're running standard tune, you're probably pegging your MAF, which means you're exceeding its reading limit. The maf reads 0-5v and there is MAF map that translates that into amount of fuel entering the engine. Traditionally the way to fix this issue is to get bigger maf housing and move the maf sensor. Once in larger housing, same amount of air will generate lower reading on the sensor so you just multiply all the air/volt map values by increase in surface area of cross section between old and new maf and you're done.

Maestro has maps for both V8 MAF with bosch element and 90mm Ford Lightning housing with Hitachi element as well as Audi TT MAF housings/MAFs. Should help mitigate some issues you're having.

Thanks for the quick reply.

I'm running a stock TT25 MAF.  Brand new.  My MAF readings are not pegged out.  I do have a 3.5" MAF that I could move to, but I don't think the flowmeter is the issue.

Whenever the boost hits 2500+ mbar i'm getting massive (+) O2 correction.  In some cases I'm getting boost spikes at 800 kg/hr which is nowhere near the limits of the TT225.  Once the values come back down below 2400 fueling settles out.  All the way up past 1000 kh/hr of airflow.

I'm stumped....

For me, it looks like the BTS protection... try to play with KFFDLBTS and KFDLBTS instead of KFLBTS and FBSTABGM.

there is a theorical EGT map (can

Log calculated EGT.

When I say O2 correction I don't mean that the ECU is dumping fuel to protect itself.  I mean that it's correcting in order to reach desired AFR.

Have a look at this log and you can see what I mean.  There are a number of short pulls with different boost levels.

notice what happens when boost exceeds the MAP limit.  I just switched to larger injectors, so things are still a little rough.

https://spreadsheets.google.com/ccc?key=0Aoetxfc9e1KpdHZ3SFpGUEwxeHVQQndPQlFDYmxib0E&hl=en&authkey=CMCs0uMM (https://spreadsheets.google.com/ccc?key=0Aoetxfc9e1KpdHZ3SFpGUEwxeHVQQndPQlFDYmxib0E&hl=en&authkey=CMCs0uMM)

 

there is something wrong, in lines 49, 50 51:
the MAF drops, TPS drops (is this TPS or the pedal position ?) and the ignition gos -3

That was me letting off the accelerator pedal.

O2s are ignored openloop (desired lambda!=1)

Not so on my ECU.

O2 correction is always used.  From idle to WOT.  The beauty of Wideband ME7.5

I don't think that your MAP limit has anything to do with the situation.

I think that your problem is with improperly scaled MAF map. You're running lean anywhere over 3.5k while at WOT. Your actual lambda keeps creeping towards 1.0 and ECU is eventually trying to correct it by increasing fuel corrections. You running lean happens way before you hit the MAP limit.

It almost seems to me like you put in larger maf housing without properly scaling the MAP. The ECU thinks it sees much less air than there is actually entering the system and it goes leaner and leaner.

How are your block 032 fuel trims? I think you will find some hair rising values in there.

If your MAF map properly scaled for TT housing.... or your injector constant proper for the injector size? Either one or both is out of whack.

Well I am using a brand new, off the shelf TT225 MAF sensor and housing.

I selected the TT225 MAF profile in the Eurodyne software, so unless the MAF linearization curve is incorrect in the software, I can't see how else it would be mis-reading.


BTW my 032 blocks are within 1% of zero.

Check main fuel correction and injector compensation tables. In my experience, they might have weird values in them.

What do you mean "weird values"?

Are you working with Maestro also?

yes.

Default maps has some values which are not normal for stock ECU in there. I would advise resetting all of these since they are very custom to some car/injector setup tapp programmed in the past.

The 660cc map tapp distributes for S4 for example has lots of custom stuff in there which causes the car to run at like 10.5 AFR up top, whatever lambda it is (0.68 or something). There are main fuel corrections and injector corrections which are much higher than needed for properly tuned car. Now S4 doesn't have wideband so it just runs like that but if he sneaked something like that into default tune for your car, it would cause havok with your  wideband O2 corrections.

For S4, This is pig rich which works since the car is fast but not as fast as when you lean it out fto 11.8 without meth or 11.4 or so with meth.

Well I have very little confidence in the base file.  When I was running the 630cc injectors that the file was programmed for my car ran like shit.  It was unacceptable for a file like that to be considered a "base file"/starting point.

I have gone through and reset many of the maps back to 1.001 and tried tuning that way.  I even went so far as to put my stock injectors back in to tune my brand new TT225 MAF with KFKHFM.

I just got back from trying to tune my new Bosch Ev14 750cc injectors and had ZERO luck with them.  I'm getting very frustrated.  I really believe that the MAF linearization map that Tapp has in there for the TT225 MAF is incorrect.  I'm trying to find a WinOLS file running the same MAF so I can compare the values in MLHFM.

For the time being I'm basically stuck at 15PSI because I have can't seem to get fueling under control.  How do I even know that the values in Maestro are even using the correct factors and offsets?  I just don't trust anything in there......I'm essentially tuning blind.

For instance, I asked for the FRLFSDP (Injection correction during RLFS) to be added and sure enough he did, however, I have always seen the pressures listed as:

-1200
-1000
-800
-600
-400
-200
0
200
400
600
800

his values are:
-300
-250
-200
-150
-100
-50
0
50
100
150
200


It seems like they're off by a scale of 4x.  Lol...  what would happen if I went and changed them to the values I think they're supposed to be?   ............BAD THINGS.  That's what would happen.

IF someone could post up a MLHFM for a TT225 MAF I would REAAAAAAAAAALLLY appreciate it.

I have an XDF file, but tunerpro crashes trying to open a map that large.

I have a few bins in winols, but haven't a clue if i have the right address.

I had a problem with MAF map when trying to open it in Tunerpro as well for S4 "M" XDF. Either newest TunerPro can finally open it or the newest revision of "M" XDF is finally allowing the program to do it, either one. Try newest TunerPro and if that still doesn't open the MAF map, look at "8D0907551M-20110122.xdf" posted on this site and move the definition for MAF map to your TT225 file. This should allow you to open and view the values.

As to your tuning gripes... you nee to find the baseline, you must be absolutely sure that you either have an accurate injector constant or maf map. Keep in mind that the Maestro injector constants are for 3 bar FPR. Most of Audis come with 4bar fpr from factory so your fueling will be off from start.


Thanks.

I am not using an S4 MAF.  I am using an TT225 MAF.  They are different diameters so the response curve will not work.

From what I understand the actual sensor is the same though.  F 00C 2G2 047

I took some time last night and plotted out every 10th value from Maestro7's "TT225" MAF map against an S3 MAF I have in WinOLS.  I attached the results below.


As for getting the injectors right, I've been using the Bosch document to calculate my injector constant and then tuning it using LTFT's.  (but maybe that was all messed up becuase my MAF profile is incorrect???)  

I haven't trusted the injector constants in Maestro for a LONG time now, they never worked for me.

750cc @ 43.5 PSI = 866CC @ 58 PSI

866cc/min x 0.684g/ml = 592.34 g/min

KRKTE = 50.2624 * (1.781/4) / 592.34 = 0.0378ms/%

The ECU doesn't seem too happy with the exact number out of the calculation.  It's close though, but I assume the differences are due to actual flow of injectors vs. rated flow.......or quite possible the completely F'd MAF curve.

I don't know what the ACTUAL Qstat of my EV14 750's is though.  Maybe I'll try and get that info out of 034 Motorsports.



 

Absolutely not true !!! ME7 cars run closed loop fueling all the time !!!

MmmBoost, what about your fuel system ? AMB engine seems to have returnless fuel system, so how you set it for your injectors and still have needed flow ? What fuel pump you use ? If you max out your fuel pump on WOT and high rpms then you will have exactly that i.e high possitive 02 correction because ECU start to compensate for needed fuel......

@Kelesha

I have considered that it might be my fuel system actually.  I am still running the stock returnless fuel system with the stock pump.  One of the things that makes me question that theory though is that even though my O2 corrections are high, my Injector Pulse Widths do not get excessively long.......which would indicate the ECU is compensating for lack of pressure.  Correct?

No way to run returnless fuel system, stock pump and GT2871R on 1.5bar........forget that !!!

ECU try to compensate till actual lambda meet request value on vcds block 31, and if the fuel is not sufficient for programed period of time i.e in example if for 3sec request lambda is not equal to actual lambda then ECU shut the throttle to safe the engine........

A friend of mine is running a T3/T4 which is larger than a GT2871R on his stock fuel system with no issues.  Maybe my pump is on it's way out.....

I did an experiment.  I went into KFKL and bumped up the 191.25 column to 1.5 from 3000-7000 rpms.  I went for a run at about 25 PSI and I was getting negative corrections except for the few RPMs where I spiked to about 27 or 28 PSI.

If I were to upgrade my pump and switch out my fuel system for the return system on the AEB engine, all I would have to do is change FRLFSDP to 1.000 across the table correct?

It sounds to me like you're shooting there in the dark. You have many unknown variables and you don't know what's going on.

My platform doesn't have wideband O2 so as a first investment into the tuning world I bought zeitronix with O2, 35psi boost sensor and fule pressure sensor. I can log the stuff I need to ensure I am not hitting some weird situation with fueling itself.

I don't know how much fuel can your pump supply but it would be a good guess that if you're still stock you might be running beyond what you hardware can support.

Even if you got just a cheap $20 fuel gauge and rigged it to your windshield outside of car  (duct tape it outside so that you can see the readin while WOTing) it would help you to see if your fuel pressure is holding or not.

Trust me man, I feel like I've been tuning in the dark from the beginning with Maestro7.  No documentation, mislabeled tables, zero support, no explanaition as to how to adjust anything.  You just get your cable, your base file.......and tune away.   ;)  Riiiiiiiiiiiiight.

In any case, once I found some Bosch Me7 documentation, ditched my 630cc injectors, and went to Bosch 550's the car was running fantastic........up to 20 PSI or so.  After that I was maxing out the injectors.  Since I've installed the Bosch 750cc injectors I feel like I've taken 5 steps back from where I was only 3 days ago.

I think it's prudent that I upgrade my fuel pump anyway.  2x the HP on the stock pump is just asking for trouble.  I looked into renting a fuel pressure testing kit here but no luck so I'll probably just pick up a gauge from Podi since I get a few perks from being sponsored.

I'm also going to install the 83mm MAF housing that I've had laying around here for 6 months since I feel I'm reaching the limits of the TT225 sensor in the 70mm housing.  I'll be putting back in the stock injectors to try and help tune KFKHFM properly.  Any tips on tuning that map?


Also, back to a question I had earlier.  If I wanted to convert to a return style fuel system with a pressure referenced FPR after the fuel rail, would all that's necessary to be changed is FRLFSDP set to 1.000 across the table?

"No issues" ? Does your friend ever try to meassure his fuel pressure at WOT ?

Yes thats correct but its not sufficient

Nope.  But I take everything he says about tuning with a grain of salt anyway.


Could you explain what you mean by this please?

First you wrote "he have no issues" now you wrote that you dont trust him..........thats not serious really......... :)

I mean that difference between return and returnless fuel systems when we talk about ME7 software is not only FRLFSDP

He claims he has no issues.  I know he hasn't logged fuel pressure.

.......and I have seen his logs.  He is not leaning out as I am.  But he is not running a MAF so I don't know what sort of voodoo he's worked on his car.  :)


What other changes would be necessary?

The return fuel system was changed for emissions reasons.  The returnless is definitely not ideal for increased power and boost..... So I would like to change it to the ideal setup if possible.

How is that possible on cars with a narrowband sensor?!

Made some test and see ;) If you need to know how to done that exactly...... thats the way ----> disconect your O2 sensor(s), made your car to run 15:1 AFR on boost when you request in example 11:1 from software, then connect O2 sensor(s) again and made some WOT drive tests with VCDS ;)

It will go into a limp mode if you disconnect the the 02, which will limit power - i'd have to look at the block diagram to find out what/how.

That doesn't prove it runs closed loop on boost.  I'm almost certain it doesn't.  Any time it runs closed loop on the S4 it runs lambda 1.  I have disabled closes loop above 1600rpm which lets me run labda 1.1 at cruise for better economy.  On full throttle I still run lambda 0.81 or whatever I ask for. 

Rick

Are you sure ? Because i am not ;) Fault codes for O2 sensors and limp mode with power reduce are different things :)

"Disabled closed loop" ? How you done that exactly ? and why then you need lambda sensors connected at all, only to avoid fault codes ?

I have only disabled closed loop over a certain RPM.  I have left it in place at idle and low rpm as with large injectors this is where fuelling is most unstable.

Rick

If you "disable" somehow lambda controll on certain rpms, than you may havent it available on WOT, but normally ECU monitor mixture on whole rpm and load range. On narrow band ECUs that not exactly closed loop like they run for lambda 1, but anyway its some kind of protection and its still available ;)

Not exactly closed loop is like being not exactly pregnant :P

My mistake, I assumed you were using narrowband sensors and ME7.1.

Apologies!

All good.  :)  I appreciate the help anyway.

I would also like to know the differences beyond FRLFSDP that are needed for a conversion.

Thanks,
Rey

Yes your mistake ;) And i wrote "not exactly" because on WOT correction is not with such hysteresis like for lambda 1, on WOT ECU can monitor only continuous lean/rich condition, in example if you request 11:1 and run 13:1 same time ECU accept that like still rich, but try to run 15:1 or 16:1 on boost when you request 11:1 and see whats going on with lambda correction.....suddenly you will see +25%  8)

Really thats almost same like "not exactly pregnant"...........

What you see then is a plausibility ceck/fault reaction and has nothing to do with closed loop fueling whatsoever.

Agreed.  ME7.1 does not run closed loop on WOT!

Not on ME7 and narrow bands.

No, situation what i explain is exactly from ME7 with narrow bands ;)
But if you preffer to believe that on narrow band cars ECU dont monitor the mixture when request lambda different from 1, then ok i havent problem with that  8)

So is it just that on an ME7 with narrow band, when lamba is not 1, the ECU only checks that the lambda is on the correct side of stoich?

Tony,

I thought of that.  Perhaps if you run leaner than stoich on boost it may trigger a code or even limit load.  I can't see it doing anything to fuel as it won't know how much.  It isn't closed loop, and you would need a pretty major fault for it to actually matter.

Rick

Interesting. I would like to know more. I was not aware of that either.

Does it check it on both sides of lambda=1?

Ok, how ECU will see that you run "leaner than stoich on boost" then ? From primary O2 sensor(s) right ? If thats not closed loop then what is the difference ?

And are you know exactly when ECU try to limit load emergency and how ? Are you ever had such situations ?

When I refer to a closed loop control system I assume that you have a sensor with a sufficient accurancy in the range you want to control a variable.
What you described is more of a safety switch behavior.
There are some fault reactions that can be calibrated like overboosting for example but most are not even described in the Funktionsrahmen.

  

well my upgraded pump from 034 showed up today and I got it installed.

Slowly started increasing boost while logging.......and managed to hit the area I was having trouble with........and then exceed it without running lean as hell.

So it turns out my suspicions (which I were hoping weren't true for my wallet's sake) were actually true.  The car is running like an animal in this cool spring weather.

Thank you to everyone who offered suggestions and input.

You need to thanks to Bosch engineers who made your Motronic ECU, because with your fuel problems and without Motronic safety all you will have in your case is dead engine ;)

This is a very interesting topic. I had to deal with a 225 hp TT (with a gt2871) recently that after being tuned a year ago (by us and another company) began to slowly start to run very rich (10:1 and more) flatlining our wideband 02 we use on the dyno. It is a narrowband ecu. Unfortunately the customer lives a couple of states away, and I only had one solid day with the car (in addition to running the business) and a couple of hours the next morning. It would only go very rich at wot and above 5500-5700 k. He made no changes to the car over the period of time since it was on our dyno and when it came back other than a higher flowing fuel pump. I tried to help him out remotely based solely on his description of the problem and a few low sampling rate logs. When the car arrived we tested/logged everything we could think of and everything checked out, but the car would always run so rich to the point it would hesitate really bad in the upper revs. The car put on 3-5k miles in between with it running great when it left, and for his first few all day race events (plus being driven home 3 states away). We logged fuel pressure, ran it mafless, swapped mafs, swapped 02's in a brute force attempt to resolve the issue because of the time crunch. It would do this at 15 psi and at 25 psi. Sorry to post jack, but I had my suspicions of it being a hardware fault, and was told by our software tuner the ecu ignored 02 feedback at wot, so this couldn't possibly be the case..I guess this should be in another thread altogether?